The Strategic Effects of Parallel Trade ~Market stealing and wage cutting~

Transcription

1

2 The Strategic Effects of Parallel Trade ~Market stealing and wage cutting~ Arijit Mukherjee * University of Nottingham and The Leverhulme Centre for Research in Globalisation and Economic Policy, UK and Laixun Zhao ** RIEB, Kobe University, Jaan February 00 Abstract: Why do roducers often accet arallel trade in some markets such as automobiles, clothing, toys and consumer electronics? This aer identifies two new factors, viz., market stealing and union-wage cutting, which may make arallel trading beneficial to a manufacturer. Secifically, (i) under erfectly cometitive labour markets in both the home and foreign countries, arallel trade may hel a manufacturer to steal market shares from cometitors, if it is more cost efficient or sells in more markets than cometitors; and (ii) in a unionized labour market, arallel trade may hel by lowering the unionized wage. These benefits of arallel trade disaear when such factors are removed. Key Words: Firm-asymmetry; Labour Union; Parallel imort JEL Classifications: F; F3 * School of Economics, University of Nottingham, University Park, Nottingham, NG7 RD, UK; arijit.mukherjee@nottingham.ac.uk; Fax: ** Research Institute for Economics & Business, Kobe University, Kobe , Jaan; zhao@rieb.kobe-u.ac.j; Fax:

3 . Introduction A common concern of olicy makers is to rotect the roerty rights of innovators from unauthorized roduction. While there exist strict measures (e.g., atent rotection) to guard the interest of the innovators, a closely related issue, namely, arallel trade (or grey markets), which allows unauthorized sale of a roduct, is often allowed. Many countries such as Australia, New Zealand and Singaore have liberalized restrictions on arallel trade and the Euroean Union is also very active in reducing restrictions on internal arallel trade. As noted in Scherer (994), the first major cometition olicy enforcement in the EC is related to an attemted dealershi territoriality within the EC. Malueg and Schwartz (994) confirm that olicies worldwide generally suort arallel trade. 3 While the conventional belief is that arallel trade hurts manufacturers by creating cometition, anecdotal evidence suggests that arallel trade may benefit them at least in some markets. For examle, some manufacturers, while ublicly oosed to gray market sales of their roducts, rivately do little to inhibit their flow and in some instances even go so far as to encourage these transactions (Liner, 990,. 4). A reort reared for the EU commission (NERA, 999,. ) states that [s]ome arallel trade, however, seems to be beneficial to the trademark owner. In the North American automobile market, retailers selling new automobiles must sign a contract with manufacturers forbidding them to re-sell these cars in other countries. However, over 00,000 vehicles intended for the Canadian market were resold south of the border in 00 (increased from 6,000 in 996 (Automotive News, 00)), yet the no-arallel-trade clause was not enforced, thus imosing no enalties on the retailers. As mentioned in Maskus (000), arallel trade occurs when a good rotected by a atent, coyright, or trademark, having been legally urchased in one country, is exorted to another country without the authorization of the local owner of the intellectual roerty rights in the imorting country. Richardson (00) documents that restrictions on arallel trade originating outside the EU are uite ermissible. 3 In the US, arallel trade was US$7-0 billions in the mid-980s (Cesedes et al., 988), and rose u to US$0 billion more recently (Comuter Reseller News, 00). In Euroe, the volume of arallel trade varies from 5% of sales (on aliances, motorcars and consumer electronics) to almost 5% (on musical recordings, cosmetics and erfumes) (NERA, 999). The House of Commons (999) reort shows that the volume of arallel trade in the UK motorcycle market is around 5% of sales.

4 In the resent aer, we rovide two strategic reasons for rofitable arallel trading: arallel trade may benefit a manufacturer either by stealing markets from the cometitors or by reducing the unionized wage faced by the manufacturer. The model exlains why some manufacturers often accet arallel trade in markets where roduct market cometition is visible, such as in automobile, cloths, toys and consumer electronics, and when the factor market is imerfect, such as in some Euroean and U.S. markets, and also justifies olicy markers relaxed aroach towards arallel trade. First, section examines the effect of market stealing, leaving the factor (labour) market erfectly cometitive. We use a framework of two manufacturers one of whom serves both domestic and foreign markets and the other serves only the domestic market. We find that the former firm s rofit is higher under arallel trading of its roduct. Moreover, if both firms serve both markets, arallel trading can benefit the firm that is more cost efficient. Second, in Section 3 we consider labour market imerfection by assuming a unionized home-country labour market. We demonstrate that arallel trade from a foreign country to the home country reduces the unionized wage. Even though arallel trade increases cometition in the home market, this wage-reducing effect can dominate the cometition effect, making arallel trading beneficial to the manufacturer. Previous literature has generally argued that arallel trade reduces rofits of the manufacturers (Malueg and Schwartz, 994, Richardson, 00, Maskus and Chen, 004 and Hur and Riyanto, 006, etc.). There are also works which show that manufacturers can gain from arallel trade under certain conditions, such as if a country chooses both the tariff level and whether or not to allow arallel trade (Knox and Richardson, 00), or if the roduct of the arallel trader and that of the manufacturer are differentiated (Ahmadi and Yang, 000). More recently, Raff and Schmitt (007) exlain that, in the resence of demand uncertainty, arallel trade may benefit the manufacturers if the following four

5 conditions are met: the retailers must order the roduct before the realization of demand, it is costly to maintain inventories, the states of demand are different across markets, and different states of demand need to affect the uantity demanded rather than the consumers willingness to ay for the roducts. In contrast, the resent aer rovides a new rationale for rofitable arallel trading, even if the government is not an active agent, the roducts of the manufacturer and the arallel trader are homogeneous and there is no uncertainty that creates the motives for risk diversification through arallel trade. Certainly these are imortant factors that can cause arallel trading to arise. However, our focus is on the imortance of market stealing and union-wage cutting, which has been ignored in the literature. The remainder of the aer is organized as follows. Section examines the effects of roduct market cometition with a erfectly cometitive labour market. Section 3 looks into the case of a labour union. Finally, section 4 concludes.. The effects of roduct market cometition Consider a manufacturer, firm, which sells its roduct in two countries, A and B. In country A, firm sells directly to the consumers. In country B, firm sells through an indeendent exclusive distributor, I. We assume that firm offers I a contract in the form of (w, T), where w is the wholesale rice and T is a transfer ayment (franchise fee) aid u-front from I to firm. So far, the model is in line with the vertical ricing model of Maskus and Chen (004). Now we introduce market cometition, by assuming that in country A, the roduct market is a duooly with firms and cometing with homogeneous roducts. Both firms have constant marginal costs, which is normalized to zero for simlicity. To demonstrate our result in the simlest way, we start our analysis with the assumtion that firm sells in 3

6 both countries A and B, while firm sells only in country A, which might be caused by resource constraints. This assumtion will be relaxed in Subsection.3 to see how the analysis is affected if both firms serve both countries. We assume that the inverse market demand function in each country is P =. We consider the following seuence of moves. At stage, firm gives a take-it-orleave-it contract (w, T) to firm I, which either accets or rejects it. We assume that I accets the offer if it earns at least its reservation ayoff, which is normalized to zero. If I accets the contract, at stage, firms, choose their oututs simultaneously, and the resective rofits are realized. If I does not accet the contract, at stage, there is no sale in country B, and firms and choose their oututs simultaneously and their rofits are realized. We solve the game by backward induction... No arallel trade First, consider the case of no arallel trade. If I accets firm s contract, given w and T, firms, maximize the following exressions simultaneously to determine their resective oututs: Max( ) w T, (a) A A A A BI Max( ), (b) A A A A Max( w) T, (c) BI BI BI where the first subscrit denotes the country and the second one the firm. Standard calculations show that the euilibrium oututs are A A /3 and BI ( w) /. () The rofits of firms, are then 4

7 n A B /9 w( w)/ T, n /9,. n I ( w) / 4 T Since firm gives a take-it-or-leave-it offer to I, the euilibrium transfer ayment is n T ( w) / 4. Firm then determines the wholesale rice w by maximizing Max /9+w(-w)/+(-w) /4, (3) w n which gives the euilibrium wholesale rice as w 0. The reason for this result is easy to understand. Since I sells the roduct in country B as a monoolist, firm finds no reason to distort I s outut choice. Hence, firm charges the wholesale rice that creates the maximum rofit in I, and it also extracts this rofit through the transfer ayment. Therefore, under no arallel trade, the euilibrium offer made by firm to I is ( wt, ) (0, /4), and this offer will be acceted by I. The net euilibrium rofits of firms, are resectively n n n 3 / 36, /9, and 0. (4) I.. Parallel trade Next consider the game under arallel trade. As in Maskus and Chen (004), Li and Maskus (006) and many others, in this situation, I not only sells the roduct in country A, but also in country B if it is rofitable. Hence, while offering the contract to I, firm needs to internalize this ossibility. If I accets firm s contract, given w and T, firms, maximize the following exressions simultaneously to determine their oututs: Max( ) w( ) T, (5a) A A A AI A BI AI Max( ), (5b) A A A AI A 5

8 Max( w) ( w) T. (5c) BI, AI BI BI A A AI Standard calculations give the euilibrium oututs as ( w) / 4, ( w) / and ( 3 w) / 4. (6) A A The rofits of firms, are BI AI A B ( w) / 6 w(3 5 w) / 4 T,, ( w) / 6. I ( w) / 4 ( 3 w) / 6 T The euilibrium transfer ayment is obtained as T w w ( ) / 4 ( 3 ) / 6. Firm then determines the wholesale rice w by maximizing Max (+w) /6 w(3 5 w) / 4 ( w) / 4 ( 3 w) /6. (7) w The euilibrium wholesale rice is obtained as w 0. It is interesting to note that even if firm faces cometition from I in country A, it does not charge a ositive wholesale rice to I. The intuition can be understood as follows. On one hand, I is a monoolist seller in country B. On the other hand, under arallel trade, not only firm but also firm face cometition from I. These combined effects induce firm to kee I s marginal cost (which is the wholesale rice) low. The euilibrium offer made by firm to I is then ( wt, ) (0, 5 /6), which will be acceted. The net euilibrium rofits of firms, are resectively 3/8, /6, and 0. (8) Now we are in osition to state two roositions: I Proosition : The euilibrium oututs of firms and are lower with arallel trade than without. 6

9 n n Proof: The oututs of firms and are /3 under no arallel trade, which are greater than the counterarts under arallel trade, /4. A A A A Proosition shows that arallel trade reduces the oututs of both firms and. Proosition : Parallel trade increases the rofit of firm and reduces that of firm. Proof: The result follows immediately from (4) and (8). The reason for Proosition is as follows. Parallel trade enables I to steal business from both firms and in country A. However, since firm can use the transfer ayment to extract rofit from firm I while firm cannot, arallel trade in effect hels firm to gain from firm s business loss to firm I, making firm better off and firm worse off. The above analysis is based on homogenous roducts. However, the imlications of roduct differentiation are straightforward as follows. If firms and roduce imerfect substitutes, it will reduce the intensity of cometition between them. As a result, the business stealing effects under arallel trade will be weakened. At the extreme situation, if the roducts of firms and are isolated, arallel trade will not affect the market share and the rofit of firm, since the roduct of the arallel trader (which is actually the roduct of firm ) is also isolated from firm s roduct. Then arallel trade increases cometition only for firm. Hence, Proosition holds if the roducts of firms and are not too much differentiated..3. Cometition in both markets In this subsection we extend the model to cover cometition in both markets. We find that rofit raising arallel trade can still occur if the firms differ in marginal costs. 7

10 Consider that both firms and serve both markets A and B. Each firm sells directly to the consumers in country A, but firms and sell their roducts through indeendent exclusive distributors I resectively in country B. 4 However, we assume that the marginal cost of roduction is c for firm, while it has been normalized to zero for firm. Also, c / 53, which is sufficient to ensure that all euilibrium oututs are ositive. We consider the same timeline as in the revious section..3.. No arallel trade If firms and give the offers ( w, T ) and ( w, T ) to I and the offers are acceted, these firms maximize the following exressions simultaneously to determine their oututs: Max( ) w T, (9a) A A A A BI Max( c) ( w c) T, (9b) A A A A BI Max( w ) T, (9c) BI BI BI BI Max( w ) T. (9d) BI BI BI BI Standard calculations show that the euilibrium oututs are ( c) / 3, A ( c) / 3, ( w w) / 3 and ( w w) / 3. BI The rofits of firms,, I are then BI A n A B ( c) / 9 w( w w) / 3 T, n A B ( c) / 9 ( w c)( w w) / 3 T, n I ( w w) / 9 T, n I ( w w) / 9 T. 4 In a different context, Ziss (997) considers the effects of exorting by the manufacturers through distributors on strategic trade olicies. 8

11 Since firms and give take-it-or-leave-it offers to resectively I, the euilibrium transfer ayments are T w w and n ( ) / 9 T w w. Firms and n ( ) / 9 determine the wholesale rices w and w resectively by maximizing Max (+c) /9+w (-w +w )/3+(-w +w ) /9, (0a) w Max (-c) /9+(w -c)(-w +w )/3+(-w +w ) /9, (0b) w n c which gives the euilibrium wholesale rices as w and w 5 n 8c. Hence, 5 both firms and charge wholesale rices which are lower than their marginal costs of roduction. It is clear that the business stealing motive in a Cournot oligooly is the reason for this tye of ricing strategies. Firms and want to reduce the marginal costs of I resectively to make them more cometitive in country B, and then use the transfer ayment to extract the gain from higher cometitiveness. This logic is in line with the incentives for strategic searation as in Vickers (985) and Ziss (997). Since c / 53, we obtain ositive euilibrium oututs, evaluated at the euilibrium wholesale rices. Under no arallel trade, the euilibrium offers made by firms and to I are resectively ( n n w, T ) ( ( c) / 5, 4( c) /5) and ( n n w, T ) ( ( 8 c) / 5, 4( 3 c) /5), and these offers will be acceted by I. The net euilibrium rofits of firms, I are resectively n n n n (43 c 97 c ) / 5, (43 08c 6 c ) / 5 and 0. () I I 9

12 .3.. Parallel trade Next consider the game under arallel trade. Then I not only sell in country A, but also in country B if that is rofitable. Hence, when offering the contracts to I, firms and need to internalize this ossibility. If I accet the contracts of firms and, the euilibrium oututs are determined by maximizing the following exressions: Max( ) w ( ) T, (a) A A A AI AI A BI AI Max( c) ( w c)( ) T, (b) A A A AI AI A BI AI Max ( w ) ( w ) T, (c), BI AI BI BI BI A A AI AI AI Max ( w ) ( w ) T. (d), BI AI BI BI BI A A AI AI AI Standard calculations yield the euilibrium oututs as A ( c w w) / 5, A ( 4 c w w) / 5, BI ( w w) / 3, BI ( w w) / 3, AI ( c 4 w w) / 5 and AI ( c 4 w w) / 5. The rofits of firms,, I are then A B ( c w w) /5 w[( w w)/3 ( c 4 w w)/5] T, ( 4 c w w) / 5 ( w c)[( w w) / 3 ( c 4 w w) / 5] T A B, I BI AI ( w w) / 9 ( c 4 w w) / 5 T, I BI AI ( w w) / 9 ( c 4 w w) / 5 T. The euilibrium transfer ayments are T w w c w w ( ) / 9 ( 4 ) / 5 and T w w c w w. The euilibrium wholesale rices can be ( ) / 9 ( 4 ) / 5 found as w (34 77 c) /88 and w (34 99 c) /88. It is clear that the wholesale 0

13 rices are higher under arallel trade than under no arallel trade. The total euilibrium rofits of firms, I are resectively ( c 3867 c ) / 448, ( c 860 c ) / 448, 0. (3) I I Proosition 3: Parallel trade reduces the rofit of firm but increases that of firm if * c c arox / 50 (.). Proof: We comare the rofits of firm under arallel trade and under no arallel trade. n Secifically, using the exressions in () and (3), we have that ( )/ c 0 for n n c [0, / 53]. Further, at c / 53, but at c 0. Hence, there exists c * / 50 ( arox.) such that the rofit of firm is higher under arallel trade than under no arallel trade if * c c, which roves the result. Some exlanations are in order. Under arallel trade, wholesale rices affect the intensity of cometition in both countries A and B. A lower wholesale rice charged by a firm, say firm, not only reduces the rofits of firm in both countries, it also tends to reduce the rofit of firm in country A. This negative imact induces the firm to charge a relatively higher wholesale rice under arallel trade comared with no arallel trade. Even if the roducts of both firms and are being traded in a arallel way, the rent extracting effect is stronger for the more cost efficient firm. If firm is very much cost efficient than firm, arallel trade makes the former firm better off by extracting a significant amount of rent from the latter firm. In contrast, if we remove the marginal cost difference between the two firms, their rent shifting effects offset each other, and as a conseuence, the intensity of roduct market cometition will be increased, reducing the rofits of both firms.

14 3. Unionized labour market in the home country In this section we examine the role of unionized labour markets. To isolate its imacts, we make some changes to the above analysis by assuming that firm is a monoolist roducer of the roduct. This firm has a lant in both country A (the home country) and country B (the foreign country), and serves both countries from the resective lants. We assume that the labour market in country A is unionized, while that in country B is erfectly cometitive, and the reservation wage rates in both countries are c, which are assume to be zero for simlicity. There is a transortation cost t for trade between the countries. We assume that the inverse demand from the consumers in country A is P while it is in country B is PB a B, with a <. This structure is similar to the artial FDI case of Lommerud et al. (003), who exclude arallel trade. We adot the right-to-manage 6 model of labour unions, 7 where the labour union chooses the wage rate and firm, which reuires one unit of labour to roduce one unit of outut, determines the emloyment/outut level. We consider full bargaining ower of the labour union to demonstrate our results in the simlest way. A A 3.. No arallel trade We analyze the following moves of the game. At stage, the labour union sets the wage in country A. At stage, firm hires labour and roduces in countries A and B. Then rofits are realized. The game is solved by backward induction. Given the wage rates in both countries, firm maximizes: 6 We refer to Vannini and Bughin (000), Lommerud et al. (003), Lóez and Naylor (004), Skaksen (004) and Mukherjee (007), to name a few, for works on the right-to-manage model of labour unions. 7 The efficient bargaining model, which stiulates that the firms and unions bargain over wages and emloyment, is an alternative to the right-to-manage model. See, Layard et al. (99) for arguments in favour of right-to-manage models.

15 Max( w) ( ). (4), A B A A B B The euilibrium oututs in countries A and B are resectively A ( w) / and a/. (5) B Therefore, the labour demand in country A is L ( w) /. The labour union maximizes the following exression to determine the wage rate: w( w) Max, (6) w which gives *, n w /. (7) We assume that t w *,n, i.e., t /, (8) which ensures that firm has no incentive to exort its roduct between the countries, given that we have assumed away the cost of setting u a business in country B. Even if the setu cost is ositive but small enough, firm refers to roduce in country B. The total rofit of firm is n ( w) a 4a A B. (9) Parallel trade Assume that there is a firm in country, called K, who buys the roduct of firm in country B and sells it back to country A. It is assumed that firm does not know the identity of the arallel trader as in Hur and Riyanto (006), 8 and as in the revious section the arallel trader derives utility only from rofit but not from consumtion. 8 If the arallel trader can be identified, firm will not sell the roduct to the arallel trader if arallel trade makes it worse off. Here we show that arallel trade benefits firm, and thus it has less incentive to restrict arallel trading even if it knows the identity of the arallel trader. 3

16 Hence, the demand in country B now comes from both consumers and the arallel trader. While roducing its outut in country B, firm needs to internalize this effect. We consider the following game under arallel trade. At stage, the labour union charges wage in country A. Then at stage, firm makes its roduction decisions in countries A and B, and K buys the roduct of firm in country B and sells it in country A. Then the rofits are realized. The game is still solved by backward induction. Since K buys the roduct in country B, it creates demand in that country along with the demand from the consumers. However, the demand by K deends on its sales in country A. Hence, while determining the outut in country B, firm should correctly anticiate the demand from K and adjust the demand function in country B accordingly. Given the wage rate in country A, the transortation cost t and the rice P B at which K buys the roduct in country B, K determines its outut to maximize: Max( P t) (0) AK A AK B AK and firm maximizes the following exression to determine outut in country A: Max( w). () A A AK A The euilibrium oututs of K and firm in country A are resectively AK ( PB t) w w PB t and A. Firm anticiates this demand from K. 3 3 Hence, the total inverse demand in country B can be exressed as P B 3a t w 3 5 B, which firm should correctly anticiate and use in determining the outut in country B. 9 Therefore, the oututs AK, A and B are determined by maximizing the exressions (0), () and the following exression: 9 In contrast to the last section, where the demand from the distributors affects the derived demand for firm s roduct, here the demand from the arallel trader affects the market demand. 4

17 3a t w 3 Max B 5 B ( ) B. () The euilibrium oututs and the rice in country B at a given w are resectively AK ( PB t) w w PB t 3a t w, A, B and P B 3a t w. 0 The wage in country A is then determined by maximizing the following exression: Max w w( w PB t). (3) 3 The euilibrium wage is PB t w 4. Given P B 3a t w 0, we obtain the euilibrium wage in country B to be *, 3a 8t w. 39 Given the euilibrium wage, the euilibrium oututs and the rice in country B are resectively AK 40 a 56t 6a 6t 5 60a 35t, A 0, B and P B 5 60a 35t. It can also be found that the demand by the consumers in country B at 95 P B is 35a 5 35t BC a PB. Note that BC AK B. 95 We assume that 40 a t, which ensures that AK 0 and B 0. Further, if 56 3 a (.06, arox.), then our assumtion (8) imlies that BC 0. Hence, we assume 54 in the following analysis that 40 a t, where the first ineuality comes from (8). 56 However, the range of t over 40 a [, ] 56 is non-emty if 4 a.57( arox.). We 7 restrict our attention to.06 a.57. Also incentive to exort its roduct between the countries. t *, w for.06 a.57, i.e., firm has no 5

18 Proosition 4: The euilibrium wage in country A is lower and firm s euilibrium outut is higher in country A under arallel trade than under no arallel trade. Proof: We find that w w if *, n *, 7 6a t, which holds for 6 40 a t. 56 Firm s outut in country A is higher under arallel trade than under no arallel trade, i.e., 6a 6t if a t, which holds for a t. 56 The intuition for the wage reducing effect of arallel trade is as follows. Parallel trade increases cometition in country A. Hence, ceteris aribus, it reduces the labour demand faced by the labour union in the country, which in turn forces the union to reduce its wage hike. On the one hand, the increased cometition due to arallel trade tends to reduce the outut of firm in country A; on the other hand, the lowered wage tends to increase firm s outut. Our result shows that the wage-reducing effect dominates the cometition effect, and thus increases firm s outut in country A. The total rofit of firm under arallel trade is ( 6a 6 t) (5 60a 35 t) A B. (4) Proosition 5: The rofit of firm is higher under arallel trade than under no arallel trade for 40 a t and.06 a Proof: We have 0 t for 40 a t. The comarison of (9) and (4) at 56 n t / shows that for.06 a.57. 6

19 The reason for the above result is as follows. Parallel trade intensifies cometition in country A, though it creates higher demand in country B. For a given wage, the cometition effect dominates the demand-raising effect, and thus arallel trade tends to reduce firm s rofit. This is clearly true if the labour markets in both countries are erfectly cometitive (see Aendix for detailed roof). However, one must take into account another effect. As shown in Proosition 4, arallel trade reduces the union wage in country, and therefore, hels to reduce firm s marginal cost of roduction in country A. The beneficial wagereducing effect (along with the demand-raising effect) outweighs the negative cometition effect and hence increases firm s rofit under arallel trade. That is, in the resence of a labour union, arallel trade may benefit a manufacturer by reducing the unionized wage. To show the role of unions in the simlest way, we have given the labour union full bargaining ower. The other extreme case is full bargaining ower of the manufacturer. In this situation, the euilibrium wage is eual to the reservation wage of the worker. The analysis under this situation is similar to that shown in the Aendix. Hence, it is clear that the resence of a labour union may make a manufacturer better off under arallel trade if the bargaining ower of the labour union is not very low. The market demand function in country A, from which the labour demand is derived, lays an imortant role in determining the beneficial effects of arallel trade. We may conjecture the case of a different demand situation. For examle, as the elasticity of the market demand function in country A increases, it will have a lower adverse imact on firm s roduction and labour demand in that country following arallel trade, since a relatively smaller reduction in the rice of the final goods is reuired to accommodate the outut of the arallel trader. Hence, as the elasticity of the market demand function increases, firm needs to reduce its outut by a lower amount following arallel trade. In other words, both the adverse cometition effect and the favourable wage-reducing effect 7

20 fall with a more elastic market demand function. However, the demand-raising effect in country B remains, thus increasing the ossibility of a beneficial arallel trade. 4. Conclusion It is generally believed that arallel trade reduces rofits of the manufacturers. However, emirical evidences do not always suort this view. In this aer, we identify two new factors under arallel trade, viz., market stealing and union-wage cutting, which may be resonsible for generating higher rofits to some manufacturers. Market stealing arises in situations when the manufacturer is even slightly more efficient or sells in more markets than its cometitors. In these situations, the manufacturer can strategically take advantage of arallel trading in its cometition with rivals. Similarly, it can also use this advantage in an imerfect factor market, such as when labour is unionized. As long as the union has some bargaining ower for wage hikes, arallel trading can hel the manufacturer to weaken the wage demand. For these reasons, many manufacturers often accet arallel trade in some markets. 8

21 Aendix Parallel trade makes the manufacturer worse off without a union Consider the set u of Section 3 with the excetion that the labour markets are erfectly cometitive in both countries. For simlicity, assume that the reservation wage is zero in both countries. If there is no arallel trade, standard calculations show that the total rofit of firm is n a. 4 The euilibrium values under arallel trade follows immediately from Section 3 3a 8t 3a t with w = 0. We have A 0, B, 30 6 AK 4 3a 8t, 5 BC 7a t and 0 P B 3a t 0. The boundary conditions are as follows: AK and B are ositive for a a t, and BC 0 for t 0 if 8 a, but BC 0 for 7 7a t if The total rofit of firm under arallel trade is ( 3a 8 t) 5( 3a t). The rofit of firm is convex in t, and t for 33a 9 t if 6 a 9 33, but t 0 for t 0 if 9 a. 33 The relevant values of t and a are 7a 4 3a t (, ) and 8 a, 7 4 3a t [0, ) and 8 9 a, and a 9 4 3a t (, ) and 9 a n Standard calculation shows that. 9

22 References Ahmadi, R. and R. Yang, 000, Parallel imorts: challenge from unauthorized distribution channels, Marketing Science, 9: Automotive News, 00, Gray market yields gold, 76, Issue 597, February 5. Cesedes, F., E.R. Corey and V.K. Rangan, 988, Gray markets: causes and cures, Harvard Business Review, 75-8 (July-August). Comuter Reseller News, 00, Vendors aim to suash gray market 963, Setember 7. House of Commons, 999, Trademarks, fakes and consumers, 8th Reort, Trade and Industry Committee (htt:// Hur, J. and Y. E. Riyanto, 006, Tariff olicy and exhaustion of intellectual roerty rights in the resence of arallel trade, Oxford Economic Paers, 58: Knox, D. and M. Richardson, 00, Trade olicy and arallel trade, Euroean Journal of Political Economy, 9: Layard, R., S. Nickell and R. Jackman, 99, Unemloyment, macroeconomic erformance and the labour market, Oxford University Press, Oxford. Li, C. and K. Maskus, 006, The imact of arallel trade on investments in cost-reducing research and develoment, Journal of International Economics, 68: Liner, S., 990, The legal and economic asects of gray market goods, Quoram Books, London. Lommerud, K.E., F. Meland and L. Sørgard, 003, Unionised oligooly, trade liberalisation and location choice, The Economic Journal, 3: Lóez, M.C. and Naylor, R.A., 004, The Cournot-Bertrand rofit differential: a reversal result in a differentiated duooly with wage bargaining, Euroean Economic Review, 48:

23 Malueg, D. A. and M. Schwartz, 994, Parallel trade, demand disersion, and international rice discrimination, Journal of International Economics, 37: Maskus, K., 000, Parallel trade, The World Economy, 3: Maskus, K. E. and Y. Chen, 004, Vertical rice control and arallel trade: theory and evidence, Review of International Economics, : Mukherjee, A., 008, Unionised labour market and strategic roduction decision of a multinational, The Economic Journal, 8: NERA, 999, The economic conseuences of the choice of a regime of exhaustion in the area of trademarks, Final reort for DGXV of the Euroean Commission, February. Raff, H. and N, Schmitt, 007, Why arallel trade may raise roducers rofits, Journal of International Economics, 7: Richardson, M., 00, An elementary roosition concerning arallel trade, Journal of International Economics, 56: Scherer, F. M., 994, Cometition olicies for an integrated world economy, Brookings Institution, Washington DC. Skaksen, J.R., 004, International outsourcing when labour markets are unionised, Canadian Journal of Economics, 37: Vannini, S. and Bughin, J., 000, To be (unionised) or not to be? A case for cost-raising strategies under Cournot oligooly, Euroean Economic Review, 44: Vickers, J., 985, Delegation and the theory of the firm, Economic Journal (Sulement), 95: Ziss, S., 997, Strategic trade olicy and vertical structure, Review of International Economics, 5: 4-5.